Dihedral Angle Mesh Error: a fast perception correlated distortion measure for fixed connectivity triangle meshes

Abstract

In computer graphics, triangle meshes are ubiquitous as a representation of surface models. Processing of this kind of data, such as compression or watermarking, often involves an unwanted distortion of the surface geometry. Advanced processing algorithms are continuously being proposed, aiming at improving performance (compression ratio, watermark robustness and capacity), while minimizing the introduced distortion. In most cases, the final resulting mesh is intended to be viewed by a human observer, and it is therefore necessary to minimise the amount of distortion perceived by the human visual system. However, only recently there have been studies published on subjective experiments in this field, showing that previously used objective error measures exhibit rather poor correlation with the results of subjective experiments. In this paper, we present results of our own large subjective testing aimed at human perception of triangle mesh distortion. We provide an independent confirmation of the previous result by Lavoué et al. that most current metrics perform poorly, with the exception of the MSDM/MSDM2 metrics. We propose a novel metric based on measuring the distortion of dihedral angles, which provides even higher correlation with the results of our experiments and experiments performed by other researchers. Our metric is about two orders of magnitude faster than MSDM/MSDM2, which makes it much more suitable for usage in iterative optimisation algorithms.

Perceptual Quality Metrics for 3D Meshes: Towards an Optimal Multi-attribute Computational Model

“…As reviewed in the previous section, different attributes have been used for 3D Mesh Visual Quality assessment (MVQ), curvature computations [18], dihedral angles [20], Geometric Laplacian [17], and Laplacian of Gaussian curvature [22]. It is worth noting that most of existing MVQ metrics, except those working on textured meshes, focus on using individual attributes separately.…”

“…In this regard, we deploy two databases that we created earlier which contain different categories of distorted and reference objects, as well as the corresponding mean opinion scores reflecting their qualities collected from human subjects. These two databases were made available to the public (http://liris.cnrs.fr/guillaume.lavoue/data/datasets.html) and have been used in many 3D mesh quality assessment related studies [1], [2], [18]- [20], [22]. These MOS databases are described below and illustrated in Figure 2.…”

“…For example, Karni and Gotsman [17] proposed combining the RMS geometric distance between corresponding vertices with the RMS distance of their Laplacian coordinates (which reflect a degree of smoothness of the surface). Lavoué [18] and Torkhani et al [19] proposed metrics based on local differences of curvature statistics, while Vasa and Rus [20] considered the dihedral angle differences. These metrics consider local variations of attribute values at vertex or edge level, which are then pooled into a global score.…”

“…Some authors also proposed quality assessment metrics for textured 3D mesh [23], [24], which integrate both texture and geometry information. Similar to the image quality metrics, some of these latter algorithms [19], [20], [22] integrate perceptually motivated mechanisms such as visual masking.…”

Perceptual Quality and Visual Experience Analysis for Polygon Mesh on Different Display Devices

“…Wang et al [32] proposed a fast mesh perceptual distance (FMPD) metric which considers the visual masking effect and psychometric saturation effect of the HVS. More recently, TPDM [33], DAME [34], [35] were proposed which also serve as reliable quality indicators for mesh models.…”